Performance evaluation of metaheuristic algorithms for femtosecond laser pulse optimization

We report our investigation on ultrashort laser pulse optimization using metaheuristic algorithms in a system comprising a mode-locked oscillator, a spectral phase shaper, and a highly nonlinear amplifier. We implemented and evaluated the performance of three bio-inspired metaheuristic algorithms: Grey Wolf Optimizer, Genetic Algorithm, and Crow Search Algorithm. Our comparative analysis focused on the convergence of cost functions, autocorrelation traces, and optical spectra of the optimized pulses. We also conducted repeatability tests to assess the algorithm's robustness. The results demonstrate that metaheuristic algorithms are highly effective in minimizing pulse duration, with the Crow Search Algorithm consistently achieving the lowest cost and highest repeatability. Importantly, our approach is universal and can be successfully applied to a wide range of laser systems, including mode-locked oscillators, chirped pulse amplifiers, and others. This work underscores the potential of intelligent optimization strategies in advancing ultrafast photonics.